In construction, the use of phase change materials (PCM) allows the storage/release of energy from the solar radiation and/or internal loads. The application of such materials for light weight construction (e.g. a wood house) makes it possible to improve thermal comfort in summer and reduce heating energy consumption in winter. The choice of a PCM depends deeply on the building structure, on the weather and on building use: numerical modeling is indispensable. In this paper, an experimental comparative study is described, using cubical test cells with and without PCM composite. A set of experimental data is detailed, concerning the air and wall temperatures. The results are compared with a numerical modeling and show that hysteresis must be taken into account to predict correctly the heat transfer.